EE- 203

Lectures on TTL and CMOS

Dr Sheikh Sharif Iqbal, KFUPM

 5.3.4 & 5.10: Bipolar Junction Transistor (BJT) Switches:
Reference: “Microelectronic Circuits by Sedra/Smith and EE-203 Online course, KFUPM
- In chapter 5, BJT operating in active mode is used to design Amplifier.
Now we will concentrate on the other two modes. At one extreme BJT
operates in cutoff region (BJT Î OFF or IC ≈ 0) and at other extreme
BJT operates in saturation region (BJT Î ON, maximum IC)
- If vi < 0.5 volt, the EBJ is reverse biased (R.B).
Since CBJ is also reverse biased (as Vcc is ‘+’)
Î BJT is OFF or operating in cutoff mode
Thus, IB=0 , IC=0 , IE=0 , VC=VCC

- If vi ≥ 0.7 volt, the EBJ is forward biased. If CBJ is still reverse biased
(if VC > 0.7) Î BJT is ON but in Active region (can’t be used as a switch)

- With ↑ vi ; IB ↑ resulting an ↑ VB and ↑ IC. Thus, VC will ↓ and when

VC < VB by 0.5v, CBJ will be F.B. As EBJ is still F.BÎ BJT is Saturated
Thus, VBE= 0.7, VBC= 0.5, VCEsat= 0.2 , ICsat= (VCC - VCESat) / RC and IB= IC / βSat
 BJT Switches (cont’d.):
- If due to some reason (i.e. temperature change), the base current is
reduced than the transistor will enter the active mode from the edge of
saturation. So the base current is designed to be higher than the edge of
saturation value, by a factor of 2 to 10. This is called overdrive factor.
- Example: For the circuit shown, select a value for RB so that transistor
saturates with an overdrive factor of 10. The BJT have a minimum
β=30 and VCEsat=0.2V. What is the resulted value of forced β?

- Solution:
IC = (5 – 0.2) / 1 k = 4.8 mA ; IB = IC / β = 0.16 mA ;

So, 10 IB = (VCC -VBE) / RB ; RB = 2.7 k

and βF = IC / IBF = 3
 BJT Switches Parameters:
- Internal capacitance of BJT introduces a time delay between the application
of iB and the flow of iC . The measure of BJT's switching speed depends on
its delay-time (td ,n-sec), rise-time (tr ,n-sec), turn-on-time (ton), storage-time (ts)
and turn-off-time (toff). Often larger ts constitute the limiting factor on S/W
 Chapter 10 Î 10.1 Digital Logic Circuits:
- In this slide, logic level specification of digital circuits are presented.
- Analog systems uses continuous property of electrical signals, whereas
in digital systems, electrical signals represent numbers. Typically in
binary digital circuits, it is customary to choose two predetermined DC
voltage levels to represent logic high ‘1’ or low ‘0’.
- However, in order to allow for the inevitable
component tolerances and other effects that change
the signal voltage levels, two distinct voltage ranges
are usually used instead of two voltage values

- Thus, in digital circuits, if the signal voltage lies in

the range VL1 to VL2, the signal is interpreted as logic low or '0'. But if
signal voltage lies in the range VH1 to VH2, the signal is interpreted as
logic high or '1'. Note, the undefined region voltages are never used.
10.1: Characteristics of Digital Logic circuits:
Characteristics that are essential to successfully design digital circuits are;
(a) Allowable range of input and output voltage/current at gates
10.1: Characteristics of Digital Logic circuits: Cont’d…
(b) Noise Margin (≈ expressed in volt): Noise margin is a measure of the
extent to which a logic circuit can tolerate noise or unwanted signals.

Î ERROR
10.1: Characteristics of Digital Logic circuits: (Loading rule)
(c) - Fan-in of a logic gate is the number of its inputs driving the gate.
- Fan-out: The output of a logic gate often has to drive a number of other
gate inputs. However, a gate can only supply a limited amount of current.
Thus, fan-out of a gate is the maximum number of
gates, allowed to be connected at the output of
that gate, and are reliably driven by that gate.
- For logic high input, output of TTL inverter
will be logic low. Thus, it will sink current
from the load gates. Fan-out of this gate,
specifies the max. no. of load gates that are
successfully driven by the inverter output.
- If the inverter is sinking current from only one
gate, the internal voltage drop in 'IR1' is low
and vout=VR1= 'logic low' < V OL(max)
- But if the given inverter is driving 30 gates,
'VR1' is high and vout > VOL(max) or ≠ 'logic 0‘
Î ERROR
10.1: Characteristics of Digital Logic circuits: Cont’d….
(d) Propagation delay (≈ expressed in nano-seconds): tP is a measure of time
taken to change the output state after an input is applied. Thus,
tPHL Î time taken to change from ‘1’ to ‘0’ & tPLH Î time to change from ‘0’ to ‘1’
10.1: Characteristics of Digital Logic circuits: Cont’d….
(e) Power Dissipation (≈milliwatts): is the supplied power required to operate
the gate and is an important factor in designing portable digital products.

(f) Delay power product (should be as small as possible): is a figure of merit

that describe the gates power requirement and the related switching speed.
 Evolution of Logic Gates Î DL gate Î RTL Inverters:
- Primary building block used in digital circuits are inverters:
- Simplest of logic inverter uses a voltage-controlled switch (figure below)
- In early days, the switch was made using only transistors and resistors,
constituents of the resistor transistor logic (RTL) family.

DL OR Gate RTL Inverter RTL Inverter

- If logic high input is applied to this transistor switch, the transistor

becomes saturated and the switch will be on. Consequently, the circuit
will sink current from any output device and the collector emitter
saturation voltage (VCEsat), will appear in the output => logic low state
 Improvement of Logic inverters; RTL Î DTL Î TTL :
- If logic low input is applied to this transistor, the RTL
Inver
transistor will cut off and the switch will be -ter
open. Consequently, the circuit will source
current for any output device, and due to small
voltage drop in the collector resistor, the output
will be at logic high. If no load is connected
with the output, as shown in the figure, Vcc will
appear in the output port => logic high or ‘1’
DTL
- RTL switches were later improved by adding NOR
diodes in the input stage called diode transistor Gate
logic (DTL) circuits. Figure shows DTL NOR gate
- Further improvement of the DTL gates were
achieved by using multiple transistors, called
transistor transistor logic (TTL) circuits,
- The logic gates belonging to this family were very popular due to its
microchip- area-efficient design and the use of multi-emitter transistor.
TTL Î CMOS Inverters Î BiCMOS : Active
Pull-up

- MOS is used to make voltage controlled switches,

as it consumes low power and has very small size.
- In CMOS inverters, complementary switches
Active
made of NMOS and PMOS are used as a pull- Pull-down CMOS
Inverter
down and pull-up networks, respectively.
- If Vin = logic low or ‘0’, the PMOS will be ON,
whereas the NMOS will remain OFF.
Consequently, the circuit will source current to
output device and Vout = logic high or ‘1’
- If Vin = logic high or ‘1’, the PMOS will be OFF,
whereas the NMOS will turn ON. Consequently,
the circuit will sink current from any output
device and circuit output, Vout = logic low or ‘0’
- BiCMOS combines advantages of TTL & CMOS
Another Type (ECL) of Logic Inverters :
- Emitter coupled logic (ECL) gates uses double throw switches to
realize logic inversion.
- This current steering logic arrangement is implemented using
differential amplifiers, & yields the fastest available digital logic gates.
- This is because the ECL operates BJT’s in active modes, and thus don't
have storage time delay (from operating in Saturation and cutoff regions)
Resistor Transistor Logic (RTL) Gates: (For your information only)

AND gate NAND gate OR gate

Example: Find
the logic function
implemented by
this RTL circuit.

Solution: SR Flipflop
(vin1=S, vin2=R vout=Y)
Transistor-Transistor Logic (TTL) Inverter: (using BJT Switches)
- Evolved from DTL circuits. Diodes are replaced with multi-emeter BJT’s for
more area-efficient IC desing. In TTL, BJT’s operate in inverse active mode.
- TTL has Input stage, Driver stage and Output stage. Say VCC= 5 v , VD (any)= 0.7 v,
VCEsat(any)= 0.1v, VBE(any)= 0.7 v
TTL NAND Gates: As per design, Q1 Îinverse active (CBJ is FB & EBJ is RB)

Characteristics of TTL gates

TTL Example Problem and solution:
Basic ECL (Emitter coupled logic ) Inverters :
- ECL uses two voltage level about '-0.8 v' for logic-high and '-1.8 v' for logic-low
- In Reference Stage: Circuit
components are selected to
set the value of 'VRef=-1.3v',
which is midpoint of signal
VINV logic swing from -.8 to -1.8v
-1.1 v
-0.8 v
- Input Stage: if Vin>(Vref+4VT),
‘0’
Q1 is ON and Q2 is OFF. So
-0.8 v =‘1’
current flows through Q1. But,
if Vin< (Vref+4VT), Q1 is OFF
and Q2 is ON. ‘I’ flows via Q2
- Output Stage: Emitter follower is used to shift the output voltage
logic level down by 'VBE(Q4)' (or 0.8 v) from that of 'VINV'. This
results (shown later) to an ECL output voltage logic swing of ; -0.8v to -1.8v.
This stage also provides the gate with low output resistance and large output
current to improve the gate FANOUT capability. The outputs are referenced to
GND to improve the noise immunity of the gate.
 10.2: CMOS Inverter: - MOS logic circuits dissipates much less power. The
high Zin of MOS allows temporary charge storage, applied in memory circuits. MOS
IC's has higher integration level due to optimized feature size and tight circuit packing.
For ‘0<vi<Vt’; since vGSN <Vt
QN is OFF and since vSGP>Vt
QP is ON. So QP pulls-up the
output to logic-high level.
Thus, v0 = VOH
For ‘Vt<vi<VIL’; SAT QN acts
as a load and QP remains ON.
CMOS inverter
So QP pulls-up the output to
See 10.2 (pg 957) for equations for Noise Margin, RDSn , RDSp , Vth ..
logic-high level Thus, v0=VOH
For ‘VIL<vi<VIH’; QN and QP
don’t act as switch. So this is a
forbidden region of operation
For ‘VIH<vi<VDD’; QN is ON,
QP acts as load or off. So the
output is Î v0=V0L
10.3.2: CMOS NOR Gate: Note pull-up network (PUN), Pull-down network (PDN)
Q’s are
either
PUN Saturated
or
Cut-off
or
works as
Active-
load

PMOS
PDN Î PUN
10.3.3: CMOS NAND Gate: and
NMOS
Î PDN
PUN

PDN
10.3.4/5: Use Demorgan’s theorem (as in EE 200) to design the electronic circuit.
- Figure below: vout= vin3 + vin1 vin2 Î PUN and dual vout= vin3 . (vin1+ vin2) Î PDN
- For a complex logic function: Y = A(B+CD) Î A+B(C+D). And the following
CMOS circuit is shown in figure 10.14 of pg 968. Note the design of PUN & PDN’s
Example on CMOS:
10.3.8: Transistor Sizing: For series connected MOSFET:

NMOS/PMOS, W/L selection is

(W/L)Q’s= 2n;
½ * 2n = n,
or ≈ to that of a
basic inverter

For parallel connected MOSFET

(W/L)Q’s= n/2
2*0.5n = n,
or ≈ to that of a
basic inverter

Pg 973
BICOMOS logic gates:
ƒ BiCMOS is a VLSI technology that unites Bipolar and CMOS circuits on the
same chip to combine the advantages of both logic families. Consiquently,
BiCMOS digital gates enjoys both, the low-power, high- input-impedance and
wide-noise-margin of CMOS and the high current- driving-capability and
high-speed-switching of BJT

ƒ Furthermore, since BiCMOS technology is well suited for implementing high-

performance analog circuits, realization of both analog and digital functions
on the same IC chip or 'system on a chip' becomes attainable.

ƒ BiCMOS Logic Gates are especially suitable for large capacitive loads
(greater than 0.5 pF or so) or when the logic gate has to drive a number of
other logic gates, requiring large amount of output current.

ƒ Modern BiCMOS, invented by intel, was available in the market in 1992 and
was eventually used to construct VLSI chips for personal computers.
Basic BiCOMOS logic inverter:

• BiCMOS inverter consist of a CMOS inverter (QP, QN ) and a BJT output stage (Q1, Q2)
• For logic low input => QN turns OFF, leading Q2 to remain OFF. But QP turns ON
and supplies base current to Q1. Consequently, Q1 turns ON and supplies load current
• But, Q1 turns off when vout=VDD-VBE1 => Disadvantage: VOH<VDD by VBE1
• For logic high input => QP turns OFF resulting Q1 to remain OFF. But QN turns ON,
and the supplied base current turns Q2 ON, which provides a large current path to
quickly discharge capacitive load. Since, Q2 turns off when vout=VBE2, & leads to a
disadvantage of VOL= VBE2 ≠ '0’. Improvement Î bleeder-resistors, R1 & R2 Î R1, not GNDed
Pass Transistor Logic (PTL) Circuits: Also called ‘transmission-gate logic'
• In PTL technology, logic circuits use series and/or parallel
combination of NMOS or CMOS switches to provide low
resistance paths to either VCC or GND. A PTL ‘AND gate’ is
shown in figure with logic function of; vout= vin1.vin2 .vin3 1Î 0

• NMOS and CMOS switches for implementing PTL switches:

(a) operation during low-to-high transition of vin

(b) operation during high-to-low transition of vin

For CMOS switches, if vout= '0', vinC= VDD and vin changes state from low to high
(vin=> VDD) => transistors QN and QP both conducts. Thus, iout is the sum of; iDN=
0.5kn(VDD-Vtn-vo)2 and iDP= 0.5kp(VDD-|Vtn|)2. But as vo=>VDD-Vtn, iDN=>0,
although iDP continues to charge C until vo=VDD. Thus, tPHL of CMOS is lower then
NMOS due to extra initial current, iDP.
For CMOS switch in fig. b, when vin=> 0, QN and QP interchanges roles.